Six Mantoux tuberculin skin tests with 1, 2, 5, 10, 20, and 50 units in a healthy male without side-effects – is skin reaction a linear function of tuberculin dose?
© Dimoliatis and Liaskos; licensee BioMed Central Ltd. 2008
Received: 03 June 2008
Accepted: 20 August 2008
Published: 20 August 2008
Tuberculosis remains a serious disease worldwide. Anti-tuberculosis campaigners many times face negative tuberculin skin tests after Bacille Calmette Guérin vaccination. Increasing tuberculin units might be a solution. However, is skin reaction a linear function of tuberculin dose? Are there any side-effects when higher tuberculin doses are administered?
Six simultaneous Mantoux tuberculin skin tests, using 1, 2, 5, 10, 20, and 50 tuberculin units (88 altogether) of purified protein derivative RT23 per 0.1 mL were applied in a healthy male Greek 35-years-old, with known natural Mycobacterium tuberculosis primary infection since five years. Skin indurations 72 hours later were 15, 22, 23, 19, 23, and 27 mm respectively.
No linear relation between tuberculin dose and skin reaction observed; skin reaction increased as tuberculin dose increased but with a decreasing rate, especially after 2 TUs, which seem correctly defined for detection of natural infection. No side-effects occurred.
During eighties the first author (ID), responsible for the antituberculosis campaign in Ioannina prefecture, Greece, faced negative post Bacille Calmette-Guérin (BCG) vaccination tuberculin skin tests (TST) many times, and he had to answer why. Was the BCG vaccine inadequate? Was the inoculation technique inadequate? Was the usual tuberculin dose in TSTs, recommended for detection of the natural infection, also adequate for detection of the artificial BCG-infection?
BCG contains weak bacilli; post-BCG TSTs may need increased tuberculin doses. He increased tuberculin units (TU) , supposing that the more the units are the more the reaction is. However, this is not known. In addition, are there any side-effects when higher tuberculin doses are administered?
ID experimented on himself using simultaneously six different dilutions: 1, 2, 5, 10, 20, and 50 TUs per 0.1 mL (88 TUs altogether). None of the three relevant papers [2–4], retrieved (22 May 2008) from Medline using the algorithm (Mantoux OR Imotest OR Tine OR "tuberculin skin test*" OR "tuberculin test*") AND (simultaneous* OR multiple OR "same person" OR "same individual" OR "same patient") AND (dose OR dosage OR unit* OR "TU"), had used more than 15 TUs altogether simultaneously on the same person, compared to 88 TUs in our case.
Tuberculosis remains a serious disease that kills millions worldwide [5–7], especially after human immunodeficiency virus pandemic and worldwide emerging multidrug-resistant strains ; moreover it has come back, even where we thought it was eliminated . This self-experiment has lost nothing of its value; contrarily it became evergreen again. Tuberculosis specialists, practising physicians, and immunity researchers might be interested on its results.
All vials of 1, 2, 5, 10, 20, 50 TU per 0.1 mL purified protein derivative (PPD) RT23 were provided by the Greek Pasteur Institute on the same day, preserved under identical conditions, and administered within a few days after production, long before their expiry date.
The tester was the domiciliary nurse of the Ioannina anti-tuberculosis campaign, with more than twenty years of experience. All six tests were injected lege-artis (one sharp bevelled disposable 25 G needle with a plastic tuberculin syringe per test; intradermal injection of exactly one tenth of a millilitre of PPD tuberculin; 4 mm white blister). Induration was read 72 hours later, using Sokal's technique (Figure 1).
Declining increases have also found Tzimakas et al (1980; four TSTs per person, 1, 2, 5, and 10 TUs PPD per 0.1 mL, in Greek healthy men and tuberculosis patients; the last ones had greater reactions) , Tzimakas et al (1984; 30 Greek healthy young women, of which ten were Mantoux tested with 2, ten with 5 and ten with 10 TUs PPD-RT23 per 0.1 mL) , Dimoliatis (1987; either 1 or 2 or 10 TUs PPD-RT23 per 0.1 mL Mantoux per group in 279 Greek schoolage children vaccinated with either liquid or lyophilized BCG; 6 groups) , Alcaide et al (1992; two Mantoux, 2 and 5 TUs PPD-RT23, in 2575 Spanish individuals) , Stuart et al (2000; two TSTs per person, 5 and 10 TUs PPD per 0.1 mL, in 128 Australian health care workers) , and Jentoft et al (2001; induration ≥ 15 mm in 32% and 41% of the Mantoux tests with 2 and 5 TUs PPD-RT23 respectively, in Norway) . Tzimakas et al (1980)  and Stuart et al (2000)  do not clarify the strain of PPD used; perhaps PPD-RT23 and PPD-S respectively, since PPD-RT23 is used in Greece and probably PPD-S in Australia.
The paradox of lower induration at 10 TUs comparing to 5 TUs (Figure 2) may be due to chance, technique errors, insufficient concentration in this particular vial, or unknown factor(s). We do not believe that technique errors occurred (same highly experienced tester, same time, exactly 0.1 mL/test, identical blister per test, identical vial preservation conditions). Chance is highly improbable because the phenomenon was observed again a few days later (November 1985), when 431 children, previously vaccinated with lyophilized BCG (April 1985), were Mantoux tested using the same vials and found the same pattern, in particular a mean diameter in mm 5.8 (standard deviation 4.8) in 74 children tested with 1 TU, 9.3 (4.4) in 15 with 2 TUs, 10.9 (5.7) in 88 with 5 TUs, 10.0 (5.4) in 94 with 10 TUs, 11.4 (5.8) in 78 with 20 TUs, and 14.5 (4.9) in 82 with 50 TUs . Thus the most probable explanation is that something was wrong with this particular 10 TUs batch. Tzimakas et al  and Stuart et al  did not offer details on reaction sizes and PPD-strain used respectively; thus it is impossible to check whether our decrease in 10 TUs exists also in their data.
The main thought of researcher, when he dared the import of so many TUs in his body (44 times the usual TST dose, and 5 to 6 times the highest bibliography dose), was that much more quantities of (not purified) mtb derivatives should exist within the body of the tubercular, and in addition in the lungs and not simply intradermally. The absence of any side effect vindicated this hypothesis. He was also based on the already use of denser dilutions , where he concluded that the TST reaction is not a linear function of the tuberculin dose, but rather a decreasingly increasing relation, compatible with the known in Immunology antigen-antibody curve . This hypothesis was also confirmed with this self-experiment: the skin reaction is an increasing function of the tuberculin dose but with a decreasing rate, especially after 2 TUs. It is worth noting that because of this decreasing rate the recommended dose for routine detection of the natural infection has correctly been defined to be 2 TUs PPD-RT23 per 0.1 mL: while the gain in mm of induration is big until this dose, it is negligible beyond this.
After six simultaneous Mantoux tuberculin skin tests with 1, 2, 5, 10, 20, and 50 TUs in a healthy male 35, i.e. after simultaneous injection on the same person 44 times more TUs than the usual TST dose of 2 TUs PPD-RT23, (i) no side-effects occurred; (ii) no linear relation between tuberculin dose and skin reaction observed; the later increased as tuberculin dose increased but with a decreasing rate, especially after 2 TUs, which (iii) seem correctly defined for detection of natural infection.
"Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal."
Bacille Calmette Guérin
- mtb :
Purified protein derivative
the name of the tuberculin used
Tuberculin skin test
- Dimoliatis JD: The size of the tuberculin skin reaction as a function of the tuberculin dose. Iatriki. 1987, 52: 538-542. (in Greek with an English abstract).Google Scholar
- Alcaide J, Altet MN, Batalla J, Plans P, Taberner JL, Salleras L, Romero MA: Comparative study of human cutaneous sensitivity to 2 and 5 international units of the PPD RT 23 tuberculin with Tween 80. Med Clin (Barc). 1992, 99 (14): 525-528. (in Spanish, with an English abstract).Google Scholar
- Stuart RL, Bennett N, Forbes A, Grayson ML: A paired comparison of tuberculin skin test results in health care workers using 5 TU and 10 TU tuberculin. Thorax. 2000, 55 (8): 693-695. 10.1136/thorax.55.8.693.PubMed CentralView ArticlePubMedGoogle Scholar
- Jentoft HF, Omenaas E, Eide GE, Gulsvik A: Comparing the adrenaline-Pirquet test with international PPD tuberculin tests. Respir Med. 2001, 95 (3): 205-211. 10.1053/rmed.2000.1023.View ArticlePubMedGoogle Scholar
- Murray CJL, Lopez AD: The Global Burden of Disease. 1996, World Health Organization, Harvard School of Public Health, and World Bank, 361.Google Scholar
- Ivan Roitt: Roitt's Essential Immunology. 1997, Blackwell Science, Oxford, 336,269,300., 9Google Scholar
- Lao LY, De Guia TS: Tuberculin skin testing: Determinants and reaction. Respirology. 1999, 4: 311-317. 10.1046/j.1440-1843.1999.00198.x.View ArticlePubMedGoogle Scholar
- World Health Organization, Regional Office for Europe: Health 21 – the health for all policy framework for the WHO European Region. 1999, Copenhagen, 47.Google Scholar
- Tzimakas C, Zarogoulidis K, Anastasiadou B, Dimitriadis G: Observations on the performance of tuberculin skin tests with different tuberculin units. Proceedings of the 8th Medical Congress of Greek Armed Forces. 1980, 71-(summary 113; in Greek).Google Scholar
- Tzimakas C, Kalafati-Tzimaka H, Kakavelas H, Kapaltzoglou A: Interference of simultaneous tuberculin skin test. Galenus. 1984, 26 (1): 83-88. (in Greek with an English abstract).Google Scholar
- Dimoliatis I: A vaccination program using lyophilized BCG in 1690 children aged 3 to 18; before and after vaccination tuberculin skin testing. Proceedings of the Third Greek Medical Congress on Thoracic Diseases . 1986, Thessaloniki, Greece, 119-(abstract 103; in Greek).Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.